Hepatocellular Carcinoma (HCC) is the most common liver malignancy. The molecular mechanism of the development of HCC remains poorly understood. Markers for early diagnosis, cell surface molecules as treatment targets, and molecular classification of HCC would address an important medical need. Recent evidence suggests that the phenotypic diversity of tumors is accompanied by a corresponding diversity in their gene expression patterns. Using DNA microarray with 23,000 genes, our recent data showed consistent differences between HCC and non-tumor liver tissues. Each tumor has its own distinct gene expression pattern. This gene expression pattern appears to reflect the underlying genotypic differences of each tumor. In search for a better understanding of the development of HCC, we propose a systematic characterization of the gene expression profiles in HCC. We will first establish the gene expression patterns of over 150 liver tissue samples using cDNA microarray containing 46,000 clones. Using different statistical methods, we will correlate gene expression patterns in HCC with genetic variations among the tumors, for example, p53 and beta-catenin mutations; as well as features of clinical significance, for example, vascular invasion, tumor recurrence, and patient survival. Next, we will characterize the DNA copy number variations in HCC samples using array based comparative genomic hybridization (aCGH). The parallel analysis of the DNA copy number changes and the alteration of the gene expression level will illustrate how molecular changes in HCC affect gene expression and help to identify novel oncogenes and tumor suppressor genes. Finally, we will characterize the cellular functions of two genes (PLA2G2A and MMP14) that we have identified though our previous studies whose expression level correlated with the vascular invasion in HCC. We would like to study how the expression of these two genes functions in tumor metastasis. We will express them in HCC cell lines and examine how these two genes may affect cell growth, adhesion, and migration, both in vitro and in vivo. In summary, the described projects will yield considerable insight into the development, genetics, and molecular classification of HCC and illustrate novel mechanism of tumor metastasis. It may also provide new markers or targets for HCC diagnosis and treatment. These results will ultimately benefit patients who suffer from this malignancy.